A total of 10,289 lactations, corresponding to 4967 does distributed in 35 herds, was used to estimate heritabilities, repeatabilities, and correlations among milk yield (kilograms), fat contents, and protein contents (grams/100 g of milk) with single-trait and multiple-trait animal models using repeated records. Single-trait estimates were similar to those obtained with multiple-trait analysis. Heritabilities were 0.18, 0.16, and 0.25 with single-trait analysis and 0.17, 0.14, and 0.22 with multiple-trait analysis for milk yield, fat content, and protein content, respectively. Repeatabilities were 0.39, 0.36, and 0.47 with single-trait analysis and 0.36, 0.33, and 0.41 with multiple-trait analysis for milk yield, fat content, and protein content, respectively. Genetic correlations were -0.89 between milk yield and fat content, -0.65 between milk yield and protein content, and 0.93 between fat and protein contents. Phenotypic correlations were -0.48 between milk yield and fat content, -0.47 between milk yield and protein content, and 0.54 between fat and protein contents. Heritabilities and repeatabilities were lower, and correlations were higher, than those found in the literature. Some combination of both EBV of milk yield and protein content would be a good criterion for selection to improve milk yield while maintaining fat and protein contents.